J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.37.2.162 on 1 February 1974. Downloaded from

Journal ofNeurology, Neurosurgery, anid Psychiatry, 1974, 37, 162-170

Voluntary control of arm movement in athetotic patients

PETER D. NEILSON' From the Division of , The Prince Henry Hospital, and the Schools of Medicine and Physics, University of New South Wales, Sydney, Australia

SYNOPSIS Visual tracking tests have been employed to provide a quantitative description of voluntary control of arm movement in a group of patients suffering from athetoid . Voluntary control was impaired in all patients in a characteristic manner. Maximum velocity and acceleration of arm movement were reduced to about 30-50% of their values in normal subjects and the time lag of the response to a visual stimulus was two or three times greater than in normals. Tracking transmission characteristics indicated a degree of underdamping which was not present in normal or spastic patients. This underdamping could be responsible for a low frequency (0-3-0'6 Hz)guest. Protected by copyright. transient oscillation in elbow-angle movements associated with sudden voluntary movement. The maximum frequency at which patients could produce a coherent tracking response was only 500 of that in normal subjects and the relationship between the electromyogram and muscle contraction indicated that the mechanical load on the biceps muscle was abnormal, possibly due to increased stiffness ofjoint movement caused by involuntary activity in agonist and antagonist muscles acting across the joint.

Athetosis is characterized by involuntary move- movement. The first is normal; the motor units ments and abnormal postures which involve act independently and asynchronously, fatigue is chiefly, but not exclusively, the distal muscula- minimal and the movement is smooth and pre- ture. Many authors have attempted to describe cise. The second type of movement is patho- or to measure the involuntary movements which logical. The motor units discharge synchron- are considered important for the diagnosis and ously because reflex mechanisms dominate the classification of (Hammond, 1871; muscle contraction. The movements are weak Foerster, 1921; Wilson, 1925, Herz, 1931; Hoefer and fatigue easily. and Putnam, 1946; Carpenter, 1950; Bobath and A frequently expressed view is that control of http://jnnp.bmj.com/ Bobath, 1952; Koven and Lamm, 1954; voluntary movement in athetosis is impaired Twitchell, 1961; Narabayashi et al., 1960, 1965). because of the release of postural reflexes from Less attention has been given to measurement of inhibitory control of higher centres (Bucy, 1942; voluntary control and interaction between Bobath and Bobath, 1952; Stanley-Jones, 1956; voluntary and involuntary movement in atheto- Twitchell, 1961; Gillette, 1964). Bucy (1942) sis. Most of the theories which have been pro- considered that involuntary movements were to the involuntary movements of caused by interruption of a suppressor mechan- posed explain on September 26, 2021 by athetosis imply that the ability to organize and ism causing hyperactivity of motor and pre- execute purposive movements is also impaired. motor cortex. Bobath and Bobath (1952) com- Hoefer and Putnam (1940) studied electro- mented that patients display stereotyped patterns myographic (EMG) patterns in athetotic muscles of voluntary movement which are dictated by a and decided that there are two types of voluntary combination of hypersensitive postural reflexes. The patient is unable to contract an individual I Address for reprints: Clinical Sciences Building, The Prince Henry Hospital, Little Bay, N.S.W., Australia. muscle without a reflex spread of activity to 162 J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.37.2.162 on 1 February 1974. Downloaded from

Voluntary control of arm movement in athetotic patients 163 involve other muscles. Although the patient can length follow-up servo theory for control of initiate movement, voluntarily activating the movement (Merton, 1953). If this suggestion is appropriate motor paths, the movement itself is correct, an abnormal action TSR in athetosis largely carried out by reflex action outside his should contribute to the deterioration of volun- control. Twitchell (1961) concluded that it is not tary control. A correlation between characteris- the presence of involuntary activity which im- tics of the action TSR and features of voluntary pairs the performance of voluntary movement in control could indicate which aspects of the athetosis, but rather an inability to integrate movement disorder are caused directly by lesions postural reflexes. Mettler and Stern (1962) dis- in the or other higher centres and cussed the possibility that athetoid movements which are secondary to changes in muscle tone. may be caused by forcing voluntary effort The purpose of the experiments explained below through a more diffuse circuitry than the normal is to provide an objective description of the con- corticospinal system. Such theories suggest that trol of voluntary movement about the elbow a distinctive pattern of impaired voluntary con- joint in patients suffering from athetotic cerebral trol may exist in athetosis. palsy. Voluntary effort and emotional stress aggra- vate involuntary athetoid movement (Koven and METHOD Lamm, 1954). Indeed, involuntary movements are obvious in some patients only during volun- Ten cerebral palsied patients presenting a clinical

picture of athetosis were tested using the following guest. Protected by copyright. tary activity when they often interfere with experimental procedures: performance and in extreme cases may com- pletely disrupt the intended movement. The FREE WHEEL TEST This test has been described else- physiological basis of the interaction between where (Neilson, 1972a). The right elbow angle was voluntary and involuntary movement is un- measured by a goniometer strapped to the arm and known. A necessary step in investigating the the EMG activity of the right arm biceps muscle was mechanism is to obtain an objective description recorded by surface electrodes attached 5 cm apart of the voluntary control of movement in over the muscle belly. The EMG was amplified and displayed on a 17 in. oscilloscope for continuous athetosis. monitoring to assure that the signal was not con- The tonic stretch reflex (TSR) in athetosis is taminated by movement artefact. It was also ampli- functionally reorganized during voluntary acti- fied in a Grass 5P3 preamplifier, integrated (time vity (Neilson and Andrews, 1973). Neilson constant-0 16 sec) and recorded on a model 5 (1972c) suggested that the action TSR pathways Grass 4-channel polygraph. The elbow-angle signal might provide the feedback required in the from the goniometer was recorded on the polygraph http://jnnp.bmj.com/

/LBOW-ANGLE ._. 10.-, ELBOW-ANGLE °wx 1 Se1c FIG. 1. Section ofa typical polygraph recording ofelbow-angle, and EMG and IEMG from biceps brachii muscle recorded during a free

wheel test. By inspection, the EMG can be seen on September 26, 2021 by to have a phase lead of 1800 to 270° ahead of the flexion movement. EMG

IEMG J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.37.2.162 on 1 February 1974. Downloaded from

164 Peter D. Neilson

so that a downward pen deflection of 10 mm corre- then the motion influenced the visual information sponded to a 100 extension change in joint angle. thereby closing the feedback path. Ability to perform The subject lay supine on a couch. He was instructed voluntary movement was assessed by analysis of to oscillate the right arm back and forth about a target, IEMG, and elbow-angle signals recorded on mean 900 joint angle position as rapidly as possible. the polygraph. The misalignment or error between Thus the maximum speed of voluntary arm move- target and elbow position was computed by sub- ment could be assessed and compared with that tracting the elbow-angle signal from the target measured previously by the same method in both signal. The root mean square value of the error normal subjects (Neilson, 1 972a) and spastic patients signal provided an index of tracking performance (Neilson, 1972c). A statistical description of the for each test. Correlation functions and power EMG signal and arm movement was obtained by spectra were computed to provide a statistical de- computing correlation functions and power spectra scription of the signals for each test (Jenkins and using computer programmes described previously Watts, 1968). Coherence between target and elbow- (Neilson, 1972a). The elbow-angle power spectrum angle signal was calculated for each test (Jenkins and enabled the predominant free wheeling frequency to Watts, 1968). Coherence can be interpreted in this be measured and the amount of distortion or devia- case as the proportion of elbow-angle movement at tion from a sinusoidal motion to be appraised. Phase any frequency which was correlated with the target. lag of elbow-angle movement behind EMG activity Coherence also provides an indication ofthe patient's in biceps was evaluated from the real and quadrature tracking performance because the presence of in- components of the cross power spectrum. This voluntary movement at any frequency reduces the technique has been explained elsewhere (Neilson, coherence at that frequency. Coherence between 1972c). target and IEMG signals and between IEMG and guest. Protected by copyright. elbow-angle signals was also computed. Finally the VISUAL TRACKING TEST The elbow-angle signal and patient's tracking performance on each test was the integrated EMG (IEMG) of the biceps muscle specified by computing the gain and phase frequency were recorded on the polygraph in the same manner response characteristics describing the relationship as described above. Polygraph sensitivity was between target and IEMG, IEMG and elbow-angle reduced so that a 100 change in elbow-angle corre- and target and elbow-angle signals. These character- sponded to a 5 mm deflection of the pen. The elbow- istics apply only to the coherent portions of the sig- angle signal from the goniometer was also used to nals and therefore the target to joint angle relation- control the vertical position of a horizontal line on ship describes the closed loop voluntary control of an oscilloscope screen. Sensitivity was adjusted so arm movement. The involuntary fluctuations in that a 100 elbow-angle extension movement corre- sponded to a 1 cm downward deflection on the screen. A second not so bright and slightly defocused 14 horizontal line on the screen was moved up and down in an irregular fashion by the experimenter. The 12 position of this target line was also recorded on the polygraph. A 1 cm downward deflection on the z 10 screen corresponded to a 5 mm downward deflection of the pen. The oscilloscope was positioned so that http://jnnp.bmj.com/ it could be viewed comfortably by the patient. He 6 was instructed to move the right arm so as to keep the two lines on the oscilloscope screen super- 4 imposed. This required him to perform a series of irregular arm movements at different speeds and 2 angular displacements about the elbow. The exact nature of these movements was dictated by the 1 2 3 4 5 motion of the target line which was controlled by FREQ on September 26, 2021 by the experimenter. It involved movement through the FIG. 2. A typical power spectrum of the elbow-angle full range of flexion-extension about the elbow and signal recorded during a free wheel test. The pre- a range of speeds up to and slightly exceeding the dominant peak in the spectrum and the absence of maximum speed at which the patient could track. harmonically related peaks indicates that the move- The movements were performed in a closed loop ment was a good approximation to a sinusoid. For the fashion in the sense that the patient was required to patients tested the peak was always between 15 and use visual information to organize the movement and 2 Hz. J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.37.2.162 on 1 February 1974. Downloaded from

Voluntary control of arm movement in athetotic patients 165 elbow-angle which occurred during each test were The movement was rhythmical in each case as not coherent with the target signal and so were not verified by the predominant peak in the power included in the frequency response characteristics. spectrum of the elbow-angle signal (Fig. 2). The The gain in each case was defined as the ratio of the frequency of the predominant peak varied amplitude of the output sine wave to that of the between patients in the range 1-5-2 Hz. The size input. and each of the peak also varied between patients but the Each tracking test lasted three minutes peak to peak amplitude of the movement patient was tested four times. A rest period was mean given between each test. was always 100 to 20°. The phase lag of elbow flexion movements RESULTS behind EMG activity in the biceps muscle was determined by measuring the size of the peaks at FREE WHEEL TEST Ability to free wheel the right the free wheel frequency in both the real and arm varied between athetoid patients (Fig. 1). quadrature cross-power spectra relating IEMG to elbow-angle signals (Fig. 3). After correction for the phase lag introduced by the integrating filter it was found that the EMG to elbow flexion phase lag was always 1800 to 270°. VISUAL TRACKING TEST Satisfactory polygraph 2 REAL recordings of visual tracking test data were ob- guest. Protected by copyright. tained for all patients (Fig. 4). Target, IEMG, elbow-angle, and error signals were described statistically by power spectral curves (Fig. 5). In spite of variations between patients, a number of spectral features were consistent for all tests. FREQ. The power spectra all contained a large pre-

-10 -8 -6 -4 -2 0 dominant peak at low frequencies (<05 Hz) indicating that most tracking movements were in REAL; the low frequency peak in -2 this range. The size of ARGAND -2 the elbow-angle spectrum was always less than DIAGRAM 4 -4[ half that in the spectrum of the target signal. Ibz o The target signal, which was controlled by the experimenter, contained no frequencies higher than about 1-5 Hz as indicated by the power spectrum reaching a negligible value for all fre- quencies greater than 1-2 Hz. The elbow-angle FREQ. did not reach http://jnnp.bmj.com/ 1 2 3 4 5 spectrum, on the other hand, values until the higher frequency of FIG. 3. An illustration ofthe calculation ofphase lag negligible of elbow flexion movement behind the EMG of the 2-3 Hz because of high frequency involuntary biceps muscle for signals recorded during a free wheel movement. The power spectrum of the IEMG test. The real component (a) and the quadrature or signal was spread across the entire band (0-5 Hz) imaginary component (b) of the cross power spectrum and displayed secondary peaks between 1 and between the IEMG and elbow-angle signals are com- 4 Hz. puted. The magnitude ofthe peaks at thefree wheeling The proportion of the total arm movement on September 26, 2021 by frequency, 2 Hz in this case, are measured andplotted variance at any frequency which was correlated on an argand diagram (c). The phase lag angle 0 of with the target signal and therefore represented flexion movement behind the IEMG signal can be movement, was measured by the co- measuredfrom the diagram. The phase lag introduced voluntary into the IEMG signal by the integratingfilter is added herence between target and elbow-angle signals to to obtain the phase lag of flexion movement (Fig. 6a). For all patients the coherence started behind the EMG signal. For all patients this was in at a high value (0 6-0 8) at low frequencies (0 1- the range 1800 to 2700. 0-4 Hz) and then decreased with frequency J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.37.2.162 on 1 February 1974. Downloaded from

166 Peter D. Neilson

ELBOW- ANGLE

TARGET FIG. 4. Section of a typical polygraph tracing of elbow-angle, target, EMG, and IEMG of the biceps muscle recorded during a visual tracking test. 2 5 Hz tremor can be seen in the elbow-angle EMG signal but it is most obvious in the EMG and IEMG recordings. The target motionts were irregular. IEMG

reaching a negligible value between 1 and 1 5 Hz. but increased to a maximum value (04-08) byguest. Protected by copyright. The coherence between target and IEMG sig- 03-05 Hz. It then decreased again to a negli- nals was also computed (Fig. 6b). The IEMG gible value at a frequency between 1 and 1 5 Hz. provided a very poor measure of voluntary The coherence between IEMG and elbow- movement at low frequencies (

140 10, 10- a b ISO 1 12 b 8 :% a~~~~~~~~a T-A TI -| TARGET LU to IEMG 0

.1 ~ ~~~~~~~~8

'60

40 2 3 4 8 1 2 3 4 5 20 FREQ. FREQ.

1 2 3 4 5 1 2 3 4 b FREQ. FREQ. http://jnnp.bmj.com/

35 A70 10 _ C 30 C so d A ERROR 0 I- 2b |ANGLE Lie

2 40 2 3 4 5 FREQ. on September 26, 2021 by

2 3 4 1 2 3 4 S FIG. 6. Coherence function between (a) target and FREQ. FREQ. elbow-angle, (b) target and IEMG, and (c) IEMG and FIG. 5. Power spectra of target, IEMG, elbow-anigle, elbow-angle signals recorded during a visual tracking anid error signals recorded during a visual tracking test. Coherence provides a measure of the proportion test. In each graph power is plotted in arbitrary units ofthe variance ofthe response signal at eachfrequency equal to mnm of pen deflection squared. Frequency is which is correlated with the input signal. Frequency plotted in Hz. has been plotted in Hz. J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.37.2.162 on 1 February 1974. Downloaded from

Voluntary control of ar-m movement in athetotic patients 167

40 0 * N I-A -40 LUJ 4 Z' I -8c \'\, 4c -L -120 pendkh \. I\\ *~~~~~~~~~ \~~~~I A = _ ' "ject . N p.nduhm I- FREQ. 40I-, FIG. 8. Phase characteristics describing the mathe- -12-C \ matical relationship between IEMG and elbow-angle signals recorded during visual tracking tests. The -14 variation between individuals has been illustrated by the difference between characteristic of subject A and 00 subject B. The phase relation between force and dis- .4 . placement ofa pendulum has been graphed to illustrate .4 s..... how the phase characteristics could be altered by a change in resonant frequency. Phase is plotted in guest. Protected by copyright. degrees. Frequency is plotted in Hz on a logarithmic axis. FREQ. FIG. 7. Gain and phase frequency response curves describing the mathematical relationship between the 1 and 1-5 Hz. Since coherence between target coherent portions ofthe target and elbow-angle signals and elbow-angle signals decreased to a negligible recorded during a visual tracking test. The phase value by this frequency the concept of gain and characteristics have been approximated by those of a phase becomes meaningless for higher fre- 04 sec time delay. Gain is plotted in db and phase in quencies. The phase characteristic for frequen- degrees. Frequency is plotted in Hz on a logarithmic cies up to 1 Hz could be approximated reason- axis. ably well by the phase lag characteristic of a 0 4 sec time delay as shown in Fig. 7b. The phase characteristics between IEMG and maximum value (0 8) and then decreased with elbow-angle signals were also calculated. A sur- frequency not reaching a negligible value until prising result was the large variation between about 3-4 Hz. patients in these phase curves at low frequencies. Visual tracking performance was specified by This variation has been illustrated (Fig. 8). The http://jnnp.bmj.com/ gain and phase frequency response curves (Fig. phase lag of elbow-angle behind IEMG at 7). These curves provide a graphical description 0-1 Hz varied between patients in the range 5°- of the mathematical relationship between the 230°. target signal and the elbow-angle response. A number of features of tracking performance were similar for every patient. Gain was always less DISCUSSION than one, it increased to a maximum (0 6-0 8) at FREE WHEEL TEST The free wheeling frequency on September 26, 2021 by a frequency between 0 3 and 0*6 Hz and then of 1 5-2 Hz in athetoid patients reported here decreased to a negligible value by 1-1 5 Hz. Thus was considerably lower than the 4-6 Hz there was always a peak in the gain curve at 0 3- measured previously in normal subjects (Neilson, 0*6 Hz. Elbow-angle responses always lagged in 1972a) and overlaps the range (0-6-2 Hz) of phase behind the target signal. At low fre- spastic patients (Neilson, 1972c). The phase lag quencies (0 1 Hz) the phase lag was 0°-30°. It between EMG activity and elbow flexion move- increased with frequency reaching 1800 between ment was similar (180°-270°) for each group of J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.37.2.162 on 1 February 1974. Downloaded from

168 Peter D. Neilson subjects. As was argued previously (Neilson, (1-2 Hz), both the elbow-angle and IEMG 1972c), this large phase lag indicates that the signals contained higher frequencies apparently maximum velocity and acceleration of arm caused by involuntary athetoid activity. movement is limited by the mechanical load on In spite of involuntary activity, the athetoid the muscle, since power of the biceps muscle was patients made less low frequency movement than not clinically impaired. required for ideal tracking in which target and Inspection of the phase curves (Fig. 8) which elbow-angle signals would be identical. This is describe the relationship between IEMG and indicated by the low frequency peak in the elbow-angle changes indicates that there is great elbow-angle power spectrum always being less variation between patients. The curves are simi- than half the size of the corresponding peak in lar to the force-displacement phase curves of a the target spectrum. It was previously reported pendulum which drop through a 1800 phase that spastic patients make more low frequency range at the resonant frequency of the pendulum movement than required for ideal tracking and it (Fig. 8). The resonant frequency ofthe pendulum was suggested that the tracking test may have equals the square root of the ratio of stiffness to revealed an underlying athetoid component not inertia where stiffness is defined as the slope of apparent on clinical examination (Neilson, the static force-displacement graph. A change in 1972c). The results here suggest that this was not stiffness would therefore change the resonant the case and that excessive low frequency move- frequency at which the phase drops through ment during voluntary effort is a feature of 180°. The variation between patients in the not observed in athetosis. In athetosisguest. Protected by copyright. IEMG to elbow-angle phase curves can therefore there are low frequency involuntary movements be interpreted as a change in the resonant fre- but in spasticity there are excessive levels of in- quency of the mechanical load on the biceps appropriate voluntary movement. muscle. A reinspection of earlier results (Neilson, It seems reasonable that the amount of applied 1972a) revealed a similar insufficiency of low force required to alter the joint angle is in- frequency movement in the tracking performance creased when opposing muscle groups acting of normal subjects. An examination of the poly- across the joint contract simultaneously. The graph recordings shows that the subjects did slope of the length-tension curve of muscle has respond to almost all the low frequency target been shown to increase with contraction level changes. Perhaps the insufficiency of low fre- (Rack and Westbury, 1969; Rosenthal et al., quency power in the elbow-angle signal is (1970), implying a greater muscle stiffness. The characteristic of the way in which the central presence of involuntary activity in opposing nervous system (CNS) organizes voluntary muscle groups in athetosis will therefore increase movements. Stark et al. (1962) demonstrated the stiffness of the joint and change the mech- that in tracking a slowly changing ramp signal a anical load on the biceps muscle. Lack of series of brief movements are executed which reciprocal inhibition is an outstanding feature of cause step-like changes in position at irregular http://jnnp.bmj.com/ athetosis (Hoefer and Putnam, 1940), and intervals. The presence of intermittency in con- appears to be the main factor in increasing the trol of voluntary movement has been indicated stiffness of the joint, changing the resonant fre- by other workers (Bekey, 1962; Stark et al., quency of the mechanical load on the muscle 1962; Navas and Stark, 1968). Such inter- and reducing the maximum speed of movement mittency, as well as an attempt to correct for to about 5000 of the value in normal subjects. phase lag caused by reaction time, could be

responsible for the low frequency variance in on September 26, 2021 by VISUAL TRACKING TEST The power spectra the elbow-angle signal being less than ideal. providing statistical description oftarget, IEMG, The maximum frequency (1-I 5 Hz) at which elbow-angle, and tracking error signals each con- athetoid patients could produce a coherent tained a predominant low frequency peak (Fig. response to the visual stimulus was less than the 5). Although the target signal was controlled by maximum frequency (1 5-2 Hz) at which they the experimenter so that the fastest movements could freewheel the arm. The maximum fre- were just beyond the patient's tracking ability quency of a coherent response to a visual stimu- J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.37.2.162 on 1 February 1974. Downloaded from

Voluntary control of arm movement in athetotic patients 169 lus must therefore be limited by central processes that the CNS includes a mechanism to compen- rather than by muscle and limb mechanics. A sate for changes in mechanical load on muscle. similar conclusion was made previously for The impairment of voluntary control of arm normal subjects (Neilson, 1972a). The limitation movement measured in this study suggests that, is most likely imposed by the time required for even if it were possible to eliminate abnormal the CNS to decode sensory information and tone and involuntary movement in athetosis, the organize the appropriate motor response. The patient would still have a degree of disability due phase lag curve between target and elbow-angle to difficulty in organizing and controlling voli- response (Fig. 7) indicates that the time delay is tional movement. about 0-4 sec in athetoid patientscompared with about 0 17 sec in normal subjects. I am most grateful to the Spastic Centre of New The coherence functions (Fig. 6) demonstrate South Wales for the Centre Industries Research that the IEMG provides a very noisy measure of Scholarship which provided the financial support muscle contraction, particularly at low fre- for these investigations. I am indebted to Associate quencies (< 0 3 Hz) where about 60% of IEMG Professor J. W. Lance and Professor E. P. George fluctuations are incoherent with elbow-angle for their encouragement and advice and for pro- changes. At higher frequencies (3-5 Hz) the vision of equipment from the Schools of Medicine amplitude of the elbow-angle signal decreases and Physics. but that of the IEMG signal increases (Fig. 4). For these reasons the elbow-angle signal pro- REFERENCES guest. Protected by copyright. vides the best measure of low speed movements Bekey, G. A. (1962). The human operator as a sampled-data (<0 5 Hz) while the IEMG provides the best system. I.R.E. Transactions. Human Factors in Electronics, 3, 43-51. measure of high frequency activity (>0-5 Hz). Bobath, K., and Bobath, B. (1952). A treatment of cerebral Apart from the disrupting influence of in- palsy based on the analysis of the patient's motor be- voluntary activity the voluntary control of move- haviour. British Journal ofPhysical Medicine, 15, 107-117. Bucy, P. C. (1942). The neural mechanisms of athetosis and ment in athetosis is impaired in a characteristic tremor. Journal of Neuropathology, 1, 224-239. fashion. This is indicated by the gain and phase Carpenter, M. B. (1950). Athetosis and the basal ganglia. frequency response curves (Fig. 7). The band- Review of the literature and study of forty-two cases. Archives of Neurology and Psychiatry, 63, 875-901. width is reduced to about half and the phase lag Foerster, 0. (1921). Zur Analyse und Pathophysiologie der is more than double its value at any frequency in striaren Bewegungsstorungen. Zeitschrift fiUr die gesamte Neurologie und Psychiatrie, 73, 1-169. a normal subject. There is also a peak in the gain Gillette, H. E. (1964). Recovery of motion following cerebral curve at 03-0-6 Hz (Fig. 7) which is not present insult: an appraisal of current methods of management. for normal subjects or spastic patients. This peak Archives ofPhysical Medicine, 45, 167-176. Hammond, W. A. (1871). A Treatise on Diseases of the indicates that the voluntary control system is Nervous System, pp. 654-662. Appleton: New York. underdamped and suggests that sudden move- Herz, E. (1931). Die amyostatischen Unruheerscheinungen. ment should excite transient oscillations of the Klinischkinematographische Analyse ihrer Kennzeichen und Begleiterscheinungen. Jahrbuch fur Psychologie und arm at the frequency of the peak (03-0-6 Hz). Neurologie (Lpz.), 43, 3-182. http://jnnp.bmj.com/ The irregularities at this frequency in the error Hoefer, P. F. A., and Putnam, T. J. (1940). Action potentials of muscles in athetosis and Sydenham's . Archives spectrum (Fig. 5) and the coherence function of Neurology and Psychiatry, 44, 517-531. (Fig. 6a) support this suggestion. This finding Jenkins, G. M., and Watts, D. G. (1968). Spectral Analysis predicts that spectral analysis of involuntary and its Applications. Holden-Day: San Francisco. Koven, L. J., and Lamm, S. S. (1954). The athetoid syndrome athetoid movement should reveal a rhythmical in cerebral palsy. Pediatrics, 14, 181-192. component at 0-3-0-6 Hz. Merton, P. A. (1953). Speculations on the servo-control of movement. In The Spinal Cord, p. 247-260. A Ciba In spite of the large inter-patient variation in on September 26, 2021 by Foundation Symposium. Edited by G. E. W. Wolstenholme. the IEMG to elbow-angle phase characteristics, Churchill: London. only small differences were measured in the Mettler, F. A., and Stern, G. (1962). On the pathophysiology of athetosis. Journal of Nervous and Mental Disease, 135, target to elbow-angle characteristics. This indi- 138-146. cates that the timing of the neural signals dis- Narabayashi, H., Nagahata, M., Nagao, T., and Shimazu, H. patched to the muscles by the CNS must have (1965). A new classification of cerebral palsy based upon neurophysiologic considerations. Confinia Neurologica, been adjusted to compensate for the changed 25, 378-392. muscle contraction characteristics, thus implying Narabayashi, H., Shimazu, H., Fujita, Y., Shikiba, S., J Neurol Neurosurg Psychiatry: first published as 10.1136/jnnp.37.2.162 on 1 February 1974. Downloaded from

170 Peter D. Neilson

Nagao, T., and Nagahata, M. (1960). Procaine-oil-wax Rack, P. M. H., and Westbury, D. R. (1969). The effects of pallidotomy for double athetosis and spastic states in length and stimulus rate on tension in the isometric cat infantile cerebral palsy. Report of 80 cases. Neurology soleus muscle. Journal ofPhysiology, 204, 443-460. (Minneap.), 10, 61-69. Rosenthal, N. P., McKean, T. A., Roberts, W. J., and Navas, F., and Stark, L. (1968). Sampling or intermittency in Terzuolo, C. A. (1970). Frequency analysis of stretch reflex hand control system dynamics. Biophysical Journal, 8, 252- and its main subsystems in triceps surae muscles of the cat. 302. Journal of Neurophysiology, 33, 713-749. Neilson, P. D. (1972a). Speed of response or bandwidth of Stanley-Jones, D. (1956). The physiology of hemiplegia and man. voluntary system controlling elbow position in intact athetosis. Journal of Nervous and Mental Disease, 123, 452- Medical and Biological Engineering, 10, 450-459. 456. Neilson, P. D. (1972b). Frequency-response characteristics of the tonic stretch reflexes of biceps brachii muscle in Stark, L., Okabe, Y., and Willis, P. A. (1962). Sampled-data properties of the human motor coordination system. intact man. Medical and Biological Engineering, 10, 460- 472. Massachusetts Institute of Technology, Research Laboratory Neilson, P. D. (1972c). Voluntary and reflex control of the of Electronics Quarterly Progress Report No. 67, pp. 220- biceps brachii muscle in spastic-athetotic patients. Journal 223. of Neurology, Neurosurgery, and Psychiatry, 35, 589-598. Twitchell, T. E. (1961). The nature of the motor deficit in Neilson, P. D., and Andrews, C. J. (1973). A comparison of double athetosis. Archives ofPhysical Medicine, 42, 63-67. the tonic stretch reflex in athetotic patients during rest and Wilson, S. A. K. (1925). Some disorders of motility and of voluntary activity. Journal ofNeurology, Neurosurgery, and muscle tone, with special reference to the corpus striatum. Psychiatry, 36, 547-554. Lancet, 2, 215-219. guest. Protected by copyright. http://jnnp.bmj.com/ on September 26, 2021 by